62 VARIATION, DISTRIBUTION, AND EVOLUTION OF THE GENUS PARTULA. 
It would seem that while gzbba must have been very abundant about 1840, the 
peculiarly colored shells in question were excessively rare; even to-day the latter 
are scarce in the neighborhood of Agafia, from whose environs most of the earlier 
collections were secured. But if three bushels were taken in that territory now, 
there would certainly be many shells of the distinctive pattern, and many speci- 
mens of radiolata and fragilis as well. In several other parts of Guam remote from 
Agafia, and in Saipan, the type called bicolor by Pease is exceedingly abundant. 
Pilsbry gives in the Manual all of the fundamental statements in the earlier 
literature and also a full series of figures which illustrate many of the color-forms 
of this highly-variable species. Necessarily the descriptions are qualitative, save 
for the few measurements of record, and the color-types are described circum- 
stantially as they are known in the available museum collections or are figured by 
earlier authors. The material of the present research comprises all of the color 
types so distinguished by Pilsbry, together with others which greatly facilitate the 
task of analyzing the whole species into the major classes with their subsidiary 
divisions. The inter-relationships of these components present problems of a very 
interesting nature, especially because transitional series and convergent forms 
occur. Then, besides, in a clearly different category, the distributional relations 
of the distinguishable color-types require detailed description and elucidation. 
Putting aside for the time the material procured in Saipan, as this will be con- 
sidered separately in a later chapter, we may now take up the species as it was 
found in Guam. In view of the facts and discussions presented in the earlier 
chapters, it will suffice here to repeat only the general statement that the species 
varies throughout Guam in numerical abundance and in the characteristics dis- 
played by the several representative associations, and that in the latter connection 
the shells differ in their statistical characters as well as in the kinds and relative 
abundance of the component color-classes. 
From 30 of the entire series of localities, 2,404 adult and 804 immature animals 
were taken; from the former 6,053 young and eggs were dissected. The partly 
grown individuals are most indispensable for the proper understanding of the 
color-types, but the embryonic shells are not so generally useful, because many of the 
distinctive colors and color-patterns are developed only in later life. 
VARIATION IN NUMERICAL ABUNDANCE. 
The absolute and relative numbers of Partula gibba in the several local collec- 
tions have been duly recorded in the general census table given in Chapter II (table 
4). Inthe case of relative frequency, the approximation of the figures to the real 
numbers will naturally be closer in the larger collections; where only a few snails 
were taken, as in many of the localities of the Coast Central Region, the data are 
valuable mainly as indicating the actual presence of the species in such places. 
Furthermore, the relative abundance of gibba must be considered with due regard to 
the presence and absolute abundance of associated species like fragilis and raditolata. 
In the description of the last-named, all of the essential points in this connection 
were fully discussed and need not be restated; here we may review only the general 
features which are specific for gibba. 
